Re: [jhc] Initial profiling results

On Fri, Sep 4, 2009 at 2:36 PM, John Meacham wrote:

Hmmm.... I am not sure why that is leaking memory, but in order to build up the list of free meta vars, it has to traverse the entire envioronment, perhaps keeping track of the meta vars incrementally as we extend the environment will help the issue.

After typechecking, unevaluated thunks from freeMetaVarsEnv are causing us to retain 187MB of data. boxyMatch is responsible for another 19MB. Perhaps some strictness is in order here? -- Taral "Please let me know if there's any further trouble I can give you." -- Unknown